1354123 - Remove customize mode UI grid styling on window edges / behind the hamburger panel (and corresponding animations)

Status: VERIFIED FIXED

(Firefox :: Toolbars and Customization, enhancement, P1)

Description

[:Gijs (he/him)]

See https://mozilla.invisionapp.com/share/5ZAEYEW8M#/screens/225203398

We'll get rid of the blue/grey background with a white grid and the padding animation (the latter will require some JS code changes, I believe).

Comment 1

[bbell]

Mock-up has moved to here: https://mozilla.invisionapp.com/share/8HBR8ZD2B#/230606754_Customize_-_First_Run

Comment 2

[:Gijs (he/him)]

Attachment: Bug 1354123 - remove customize mode transition,

Review commit: https://reviewboard.mozilla.org/r/152850/diff/#index_header
See other reviews: https://reviewboard.mozilla.org/r/152850/

Comment 3

[:Gijs (he/him)]

This patch removes the transition and extra grid-styled padding/borders on the window, conditional on the MOZ_PHOTON_THEME ifdef which will track nightly until 57. As a result, the "CSS only" CART transition stuff gets omitted, and a cart run has 6 subtests (it has 9 right now). All the remaining tests are either a wash or substantial improvements.

I have 2 questions for jmaher:
- is the subtests disappearing going to break anything?
- is the fact that the .half subtests now log:

13:58:12     INFO -  PID 8864 | [#0] 1-customize-enter.half.TART  Cycles:25  Average:0.00  Median:0.00  stddev:0.00 (NaN%)  stddev-sans-first:0.00
13:58:12     INFO -  PID 8864 | Values: 0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0

Note in particular the stddev (NaN%) - going to break anything ?

Comment 4

[:Gijs (he/him)]

Patch is on top of bug 1354145, explicitly marking that dep.

Comment 5

[Joel Maher ( :jmaher ) (UTC -8)]

removing a subtest is no big deal- we should see a change in the summarized number, just make myself and :igoldan aware when this lands and we will copy the change here and will acknowledge this as a test change.

regarding the .half tests, that is not good if we have 0.0- I assume this is the subtest being removed?

Comment 6

[:Gijs (he/him)]

(In reply to Joel Maher ( :jmaher) from comment #5)
> removing a subtest is no big deal- we should see a change in the summarized
> number, just make myself and :igoldan aware when this lands and we will copy
> the change here and will acknowledge this as a test change.
> 
> regarding the .half tests, that is not good if we have 0.0- I assume this is
> the subtest being removed?

Effectively, yes, I'm ifdef-ing out the things that would cause that subtest to measure something, I think. How bad "not good" is this going to be? As in, do I need to write up a patch to ignore these subtests or something, or can we live with it?

Comment 7

[Joel Maher ( :jmaher ) (UTC -8)]

we can live with it, it will change the numbers slightly which is ok.  There are many other tests which report 0 values in a subtest.

Comment 8

[Jared Wein [:jaws] (please needinfo? me)]

Comment on attachment 8881751 [details]
Bug 1354123 - remove customize mode transition,

https://reviewboard.mozilla.org/r/152850/#review158500

Looks good, I expect a nice CART improvement here.

Comment 9

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/328777d96619
remove customize mode transition, r=jaws

Comment 10

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/mozilla-central/rev/328777d96619

Comment 11

[Ionuț Goldan [:igoldan]]

Some very noticeable improvements have landed! Thanks for the good work!

== Change summary for alert #7617 (as of June 29 2017 17:12 UTC) ==

Improvements:

 63%  cart summary osx-10-10 opt e10s     32.98 -> 12.05
 59%  cart summary linux64 opt e10s       21.09 -> 8.55
 59%  cart summary windows7-32 opt e10s   25.93 -> 10.59
 58%  cart summary linux64 pgo e10s       18.22 -> 7.65
 57%  cart summary windows10-64 opt e10s  21.98 -> 9.45
 57%  cart summary windows10-64 pgo e10s  18.02 -> 7.82
 56%  cart summary windows7-32 pgo e10s   18.66 -> 8.30

For up to date results, see: https://treeherder.mozilla.org/perf.html#/alerts?id=7617

Comment 12

[Tanvir Rahman]

I have reproduced this Bug with Nightly 55.0a1 (2017-04-06) on Windows 10, 64 Bit!

The bug's fix is now verified on latest Nightly 56.0a1

Build ID 	20170706060058
User Agent 	Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:56.0) Gecko/20100101 Firefox/56.0

Comment 13

[u554753]

Verified on Windows, Mac, and Ubuntu.